SANITIZED BUILD RECEIPT · builtgtm.ai/builds
Scoring the TAM · The first deliverable · 2026-06-13
We have a product a lot of people use. Inbound, outbound, product - three entry points to one motion. Five steps turn the sprawl into a priority list with a named move on every line. A rep walks in Monday knowing which accounts to work and what to say.
Three entry points to one motion - inbound, outbound, product - and reps couldn't tell them apart. One engine, one score, makes them honest.
Enrichment → Scoring → Brief. One pipeline, run weekly, against every cohort.
A five-step Signal Stack and a per-account brief. Same math across inbound, outbound, and product.
A Sales floor that opens Monday with the right list and the right opener - instead of guessing.
Inbound, outbound, product-led - three entry points, one motion. We had to make the score honest across all of them.
The challenge isn't getting more accounts in the funnel. It's that the funnel is full of accounts a rep can't tell apart. A free-trial signup looks like a cold lead. An inbound demo looks like an enterprise prospect. An outbound prospect looks like every other outbound prospect. The opener is the same generic line for all of them. Reps work the wrong accounts at the wrong moment with the wrong message, and the score we used to have couldn't separate them.
The hardest case is the free-trial user. They look like a customer because they have product usage, but they're really a buyer who hasn't been talked to. We needed one engine that could read product users, inbound demos, and outbound prospects on the same math - with different inputs for each - and rank them honestly against one another.
That's what the package below does. It turns a sprawl of free users, inbound demos, outbound prospects, and product accounts into one ranked list with a named move on every line. The list is honest because the score is honest. The score is honest because the evidence under it is real. The evidence is real because the enrichment ran first.
Enrichment fills the evidence. Scoring ranks the list. Brief hands the rep the context.
The engine runs weekly. Every cohort in the workbook - prospects, free users, hidden mid-market, churned win-backs - runs through the same three stages and comes out the other end as a ranked list with a brief on every Tier 1 and Tier 2 account.
Each step only works because the one before it did. Open any step for the deep dive.
Same engine. Same math. Different inputs. Different opening moves.
88-person sales team. Mid-market SaaS, $40M ARR. Runs HubSpot + Gong. Strong sales-motion match.
New VP Sales started 3 weeks ago. $32M Series B closed in March. Hiring 4 SDRs this quarter. Sequence vendor RFP open (Crustdata hit).
Jordan Lee (VP Sales) - verified email, LinkedIn confirmed, in role 3 weeks.
Open with the VP transition. Reference the 90-day new-leader buying window + the Series B. Lead with the follow-up game framing. Reach Jordan first.
Free workspace, 24 weekly users and growing. Logistics, mid-market.
Sequences + CRM sidebar - power users emerging. Not using the AI suite (0 of 4: Compose, Smart Send, Meeting Copilot, the fit score).
Workspace expanding fast. 2 people researching alternatives (G2 visits flagged).
Lead with their own usage, pivot to the AI gap. Enrich a sales DM first - no buying-committee contact in CRM yet.
Three lines per role. What changes Monday.
Built in Cowork. Next: Claude Console + Systems / SE for the production architecture.
The full engine - enrichment waterfall, scoring model, Brief Pipeline - was built end-to-end in Claude Cowork. That's how we got from a methodology decision to a live system with briefs in field inside a single iteration cycle. Cowork was the right environment for the build. It is not the right environment for the scale.
The next step is moving the production loop to Claude Console for the agent ecosystem already in flight, and partnering with Systems Engineering and Solutions Engineering on the architecture. The partnership turns this from a Cowork session a single person operates into a system the whole GTM org runs against.
Open the Deep Dive's "What scales next" block for the full ask - scheduling and retry semantics, persistent agent memory, multi-agent orchestration, observability and eval, write-back governance, secrets layer. The conversation worth having is what an AI-Native GTM system looks like in production.
